Support vision input for generation utils.

[ProKil]

This PR supports vision/image input for generation utils, which would allow interaction to be grounded in vision inputs. This PR relies on #116

📑 Description

Uses Langchain's ImagePromptTemplate to add image url to chat messages.

✅ Checks

• [ ] My pull request adheres to the code style of this project
• [ ] My code requires changes to the documentation
• [ ] I have updated the documentation as required
• [ ] All the tests have passed
• [ ] Branch name follows type/descript (e.g. feature/add-llm-agents)
• [ ] Ready for code review

ℹ Additional Information

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 41.66667% with 7 lines in your changes missing coverage. Please review.

Project coverage is 59.96%. Comparing base (8d9b9be) to head (f8e59c0).

@@            Coverage Diff             @@
##             main     #119      +/-   ##
==========================================
- Coverage   60.03%   59.96%   -0.07%     
==========================================
  Files          47       47              
  Lines        2402     2413      +11     
==========================================
+ Hits         1442     1447       +5     
- Misses        960      966       +6     

Files	Coverage Δ
tests/generation_utils/test_generation.py	68.42% <ø> (ø)
sotopia/generation_utils/generate.py	51.53% <41.66%> (-0.27%)	:arrow_down: